Research On Decoding Algorithm And Recognition Technology Of LDPC Codes

Low-density parity-check(LDPC)code is one of the linear block codes.It has the advantages of code length and code rate flexibility,close to Shannon limit,and parallel decoding.It has been widely used in various military and civil communication scenarios.Therefore,the related technology of LDPC code has been a hot issue of research in recent years,.The main function of LDPC codes is to improve the reliability of system,and the decoding algorithm is an important factor affecting this performance.However,the receiver does not know the encoding method and encoding parameters used by the sender in non-collaborative communication,and can not directly decode to obtain information content.They can only complete the acquisition of information by analyzing the intercepted information sequence and estimating the encoding method and encoding parameters used by the sender.Therefore,the research on LDPC decoding and recognition technology is a very meaningful work.Firstly,this paper introduces the coding and decoding theory of LDPC codes and the basic principles of its recognition technology,including the basic principle of LDPC codes,the construction of matrix,encoding algorithm,commonly used decoding algorithm and the classification and characteristics of recognition algorithms,which lays a theoretical foundation for the subsequent improvement of decoding algorithm and recognition algorithm.Secondly,aiming at the problem of bit error rate(BER)performance and complexity in LDPC decoding algorithm,this paper mainly aims to improve the BER performance at the expense of a small amount of complexity.On the basis of minimum sum(MS)decoding algorithm,approximation operation is used again to approximate the non-linear function in the derivation process as a piecewise function,which corrects the error in MS algorithm to some extent.In order to improve the convergence speed of decoding algorithm,this paper changes the scheduling mode of the algorithm,changes the original flood scheduling to hierarchical scheduling,and proposes a layered class fitting modified minimum sum(LCFMMS)decoding algorithm.The simulation results verify the superiority of the improved algorithm.Finally,for the existing LDPC codes recognition algorithms,many literature focus on improving the recognition performance under low signal to noise rate(SNR),ignoring the impact of parity-check matrix row weight on the recognition performance.This paper aims to improve the recognition performance of LDPC code parity-check matrix under high row weight without affecting the recognition performance of low row weights as much as possible.Therefore,this paper presents a tangent assisted least square(TLS)algorithm,which maps the parity-check relationship between the receiving sequence and the parity-check vectors to a partial tangent function,and processes the data obtained by least squares method,which improves the recognition performance of the parity-check matrix with high row weight,and analyses the theoretical performance and complexity of the improved algorithm.The simulation results verify the feasibility and superiority of the improved algorithm.